The Agent Skills Directory

Indirect Prompt Injection (SAFE): While the skill ingests untrusted data from external websites (which is a standard surface for indirect prompt injection), it does so as part of its primary accessibility audit purpose. There is no evidence of malicious intent or unusual capabilities beyond standard browser automation.
Ingestion points: Web content accessed via browser-devtools-cli commands in SKILL.md.
Boundary markers: Not explicitly defined, but characteristic of standard auditing tools.
Capability inventory: browser-devtools-cli functions (navigation, interaction, snapshots).
Sanitization: Not present, as is typical for audit-focused analysis tools.
Command Execution (SAFE): The skill uses browser-devtools-cli through Bash, which is an expected and restricted tool for its stated purpose.

accessibility-audit